AbstractAimHabitat fragmentation and loss are two of the most important factors driving current biodiversity decline. Nonetheless, the relationship between biodiversity and landscape patterns appears more complex than generally expected, depending on the species and communities involved. We aim to enrich knowledge concerning the relationship between plant diversity and landscape patterns along linear landscapes, such as Mediterranean coastal dunes. A dedicated buffering method considering multiple nested extents was developed for sampling linear landscapes (e.g. coastal or fluvial), which traditionally present a challenge for standard round or square sampling buffering approaches.LocationTyrrhenian coast of central Italy.MethodsBased on a database of plant community plots and functional traits from field measurements, for each plot we calculated taxonomic (TD) and functional (FD) diversity, which was further decomposed in functional evenness (FDeven) and mean trait dispersion (FDdisp). Relying on a land‐cover map, we computed a set of landscape metrics describing habitat loss, fragmentation and direct human disturbance at multiple extents around each plot. Diversity measures (TD, FD, FDeven and FDdisp) were then related to the landscape metrics at different scales via linear mixed‐effect models.ResultsOverall, the relationship between plant species diversity and landscape patterns was weak. We observed different responses of TD, FD, FDeven and FDdisp, which only emerged at fine‐medium scales. TD decreased with habitat loss and disturbance, while FD only with disturbance. FDeven decreased in more fragmented areas, while FDdisp was not affected by the selected landscape parameters.Main conclusionsLike other transitional areas, coastal strand and dune ecosystems exhibit steep gradients in biotic and environmental factors, are dynamic in location, and could be among the earliest to be affected by environmental drivers. However, the response of Mediterranean coastal dune plant diversity to habitat loss and fragmentation is weak. For these reasons, we propose that these plant communities are adapted to the ever‐changing nature of the coastal environment and consequently to changes in landscape pattern.